Genetically encoded molecular tools are having impact throughout neuroscience. However, each genetically encoded molecular tool is a feat of protein engineering, requiring extensive effort to design and optimize. One appealing thought is that it might be possible to create a small number of modular protein building blocks, which then could be organized in patterns, and achieve user-programmable functions, by being scaffolded together on strands of RNA. We recently derived from the Pumilio homology domain (PumHD) protein, 4 building blocks, each of which prefers one of the 4 bases of RNA, so that for any RNA of interest in a cell, a chain made out of the 4 building blocks in a defined sequence can be created to bind to that RNA (bringing along any protein payloads that are fused to that chain). We propose to adapt this Pumilio based (or ?Pum? for short) strategy to create RNA-scaffolded molecular tools that address two major classes of problem in neuroscience ? the targeting of gene expression to specific cell types in the mammalian brain, with high specificity, without requiring transgenic animals (Aim 1), and the ability to image multiple independent fluorescent reporters of physiological activity within individual neurons without crosstalk (Aim 2). We will assess, and optimize, these scaffolded molecular tools in mice (Aims 1, 2) and a non-genetic model organism (macaques), aiming to open up new frontiers of neuroscience experimentation (Aim 3). Our grant is a fast- paced, 4-year grant, which brings together experts in neurotechnology development (Boyden), primate systems neuroscience (Desimone), and nucleic acid technology (Adamala). We will share all tools freely with the neuroscience community, in order to broadly accelerate neuroscience progress.